Key Takeaways:
Fundstrat's Tom Lee said the artificial-intelligence trade remains in "very good shape" despite growing questions about its durability, citing a 62-percentage-point performance divergence between Ethereum and memory-chip stocks as evidence of broadening adoption.
The AI trade remains in "very good shape" even as investors question its durability, Fundstrat Head of Research Tom Lee said, pointing to a 62-percentage-point performance divergence between Ethereum and memory-chip stocks as evidence of broadening adoption.
"The AI trade is still in very good shape," Lee said on CNBC's "Power Lunch" on July 27. "What we're seeing is the market starting to price in the downstream beneficiaries, not just the infrastructure plays."
Ethereum rose about 24% over the comparison period while the Roundhill Memory ETF declined by roughly 38%, according to Fundstrat data. The ETH-to-DRAM measure strengthened by 72 points over the past month, indicating Ethereum is outperforming the basket of memory-related companies. Lee has described Ethereum as strengthening as an "AI downstream" asset — one that could benefit from applications built on AI infrastructure rather than only the initial spending on chips and data centers.
If the AI trade is broadening beyond hardware, it could shift capital flows into assets tied to AI application-layer revenue. Lee maintained his year-end S&P 500 target of 7,300, suggesting the broader equity rally remains intact even as the Federal Reserve navigates what he called a possible "inflation shock."
Memory shortages and infrastructure bottlenecks dominated the first phase of the AI investment cycle as companies rushed to acquire chips, storage and computing capacity from Nvidia Corp., Advanced Micro Devices Inc. and other suppliers. The Roundhill Memory ETF, which tracks companies tied to memory-chip production, fell 38% as those bottlenecks began to ease.
Lee has argued that Ethereum could become a financial settlement layer for AI agents, processing stablecoin payments, tokenized assets and machine-to-machine transactions. Under that thesis, Ethereum would benefit indirectly from AI adoption rather than directly from hardware spending. Fundstrat Head of Digital Asset Strategy Sean Farrell has also noted that Ethereum's role as an "AI downstream" asset could strengthen as the technology matures.
Lee has also endorsed analysis suggesting prediction markets are underestimating the odds of the CLARITY Act becoming law. Polymarket gave the legislation a 47% chance of passage at the time of the analysis, while Kalshi showed similar probabilities. Lee argued that recent restrictions preventing senators from trading on prediction markets meant informed activity was not fully reflected in those prices.
"Polymarket and Kalshi markets likely underestimate odds of passage," Lee said. "This is bullish."
Farrell wrote in a note to investors that he remained skeptical "this market fully reflects informed expectations given its relatively limited liquidity and recent restrictions preventing Senators from trading on prediction markets."
For investors, the broadening AI trade creates both opportunity and risk. If Lee's thesis is correct, companies tied to AI application-layer revenue — including Ethereum, software platforms and cloud services from Microsoft Corp. and Alphabet Inc. — could outperform hardware suppliers as the cycle matures. Nvidia, which trades at about 35 times forward earnings, could face multiple compression if capital spending shifts from chips to applications.
Lee has maintained his year-end S&P 500 target of 7,300, implying roughly 8% upside from current levels. He has also set near-term Ethereum targets of $7,000 to $9,000, with a longer-term scenario reaching $12,000 to $22,000 if Bitcoin climbs to $250,000.
This article is for informational purposes only and does not constitute investment advice.
